def run( self ):
## Read in excel file to data frame
excelFile = os.path.join( self.mainDir, self.excelFile )
table = pd.ExcelFile( excelFile )
t = table.parse( )
## Init parsed data containers
labelInfo = {}
frameInfo = {}
#Probe sample image for dims
img = pp.readInImg( os.path.join( self.dataDir, "sub01/EP02_01f/reg_img46.jpg" ) )
self.nh, self.nw = pp.imgRes( img, self.imgRes )
dims = self.nScales * self.nOrient * self.nh * self.nw
featureInfo = np.array([ ] ).reshape( 0, dims )
tf = 0
s = 0
## Iterate through files and generate labels and features
complete = True
## For each subject
for sub in os.listdir( self.dataDir ):
v = 0
## For each video
for vid in os.listdir( os.path.join( self.dataDir, sub ) ):
if vid.endswith( "avi" ): #skip avis
break
## Each line in main excel file proports to one video, get that info here
intLabelInfo, labelParams = CASMELabels.getLabelInfo( sub, vid, t, [s, v] )
f = 0
## Process each frame
for frame in os.listdir( os.path.join( self.dataDir, sub, vid ) ):
frameInfo = CASMELabels.getFrameParams( intLabelInfo, labelParams, f, frame )
frameInfo.update( intLabelInfo )
labelInfo.update( { tf : frameInfo } )
img = pp.readInImg( os.path.join( self.dataDir, sub, vid, frame ) )
img = pp.downsample( img, self.nh, self.nw )
kernels = self.gabor.generateGaborKernels( )
try:
featureInfo = np.vstack( [ featureInfo, gab.processGabor( img, kernels ) ])
except MemoryError:
#print "S%d O%d H%d W%d.csv" % (self.nScales, self.nOrient, self.nh, self.nw )
complete = False
break
f += 1
tf += 1
v += 1
s += 1
return featureInfo, labelInfo, complete
def parseLabels(self):
## Read in excel file to data frame
excelFile = os.path.join(self.mainDir, self.excelFile)
table = pd.ExcelFile(excelFile)
t = table.parse()
## Init parsed data containers
labelInfo = {}
frameInfo = {}
## Iterate through images and build label frame
tf = 0
s = 0
## Iterate through files and generate labels and features
## For each subject
for sub in os.listdir(self.dataDir):
v = 0
## For each video
for vid in os.listdir(os.path.join(self.dataDir, sub)):
if vid.endswith("avi"): #skip avis
break
## Each line in main excel file proports to one video, get that info here
intLabelInfo, labelParams = CASMELabels.getLabelInfo(
sub, vid, t, [s, v])
f = 0
## Process each frame
for frame in os.listdir(os.path.join(self.dataDir, sub, vid)):
frameInfo = CASMELabels.getFrameParams(
intLabelInfo, labelParams, f, frame)
frameInfo.update(intLabelInfo)
labelInfo.update({tf: frameInfo})
f += 1
tf += 1
v += 1
s += 1
return labelInfo
def parseLabels( self ):
## Read in excel file to data frame
excelFile = os.path.join( self.mainDir, self.excelFile )
table = pd.ExcelFile( excelFile )
t = table.parse( )
## Init parsed data containers
labelInfo = {}
frameInfo = {}
## Iterate through images and build label frame
tf = 0
s = 0
## Iterate through files and generate labels and features
## For each subject
for sub in os.listdir( self.dataDir ):
v = 0
## For each video
for vid in os.listdir( os.path.join( self.dataDir, sub ) ):
if vid.endswith( "avi" ): #skip avis
break
## Each line in main excel file proports to one video, get that info here
intLabelInfo, labelParams = CASMELabels.getLabelInfo( sub, vid, t, [s, v] )
f = 0
## Process each frame
for frame in os.listdir( os.path.join( self.dataDir, sub, vid ) ):
frameInfo = CASMELabels.getFrameParams( intLabelInfo, labelParams, f, frame )
frameInfo.update( intLabelInfo )
labelInfo.update( { tf : frameInfo } )
f += 1
tf += 1
v += 1
s += 1
return labelInfo